JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 1991 VOL. 6 527 Electrothermal Vaporization Sample Introduction System for the Analysis of Pelletized Solids by Inductively Coupled Plasma Atomic Emission Spectrometry Vassili Karanassios,* J. M. Ren and Eric D. Salint Department of Chemistry McGill University 80 1 Sherbrooke St. West Montreal Quebec H3A 2K6 Canada A new approach for solid sample introduction into a furnace for use with an inductively coupled plasma has been developed and tested with atomic emission spectrometry. A powdered sample is mixed with graphite and pressed into a pellet. The pellet is placed between the electrodes of a modified electrothermal vaporization device. A current causes rapid ohmic heating of the pellet and results in analyte vaporization. The vapour is swept into the plasma by an Ar carrier gas stream.In this preliminary report system characterization with single element oxide standards and testing with powdered botanical samples are described. The system shows considerable promise for rapid screening of botanical samples of environmental concern. Detection limits for Cd Zn and Pb of 1 3 and 0.6 ppb (300 800 and 150 pg) were obtained using the oxide standards. Keywords Electrothermal vaporization; sample introduction; inductively coupled plasma atomic emission spectrometry; powders; solids analysis Several approaches aimed at extending the analytical capability and utility of the inductively coupled plasma (ICP) by employing methods for the direct analysis (e.g. with little or no pre-treatment) of solid samples are currently being investigated.'-* Included in these are elec- trothermal vaporization (ETV)9-46 and direct sample inser- tion (DSI)47-s2 devices arc and spark v a p o r i ~ a t i o n ~ ~ - ~ ~ laser a b l a t i ~ n ~ ~ - ~ ~ electrical vaporization of thin filrnP and p o ~ d e r ~ ~ - ~ ~ and s l ~ r r y ~ ~ - ~ ~ ~ sample introduction systems.Of these ETV devices are most commonly used. Com- mercial and laboratory constructed ETV devices have been used with ICP atomic emission spectrometry (ETV-ICP- AES)9-33 and ICP mass spectrometry (ETV-ICP-MS)34-46 for the analysis of l i q ~ i d ~ - ~ ~ * ~ ~ - ~ ~ and solid sample^.^^-^^ In solid sampling ETV the material to be analysed is typically placed into a sample holder (z.e. a graphite cup or rod) in the form of a powder.In ICP spectrometry the role of an ETV sample introduc- tion system is to dry and/or ash (char) a sample and to generate analyte vapour. The vapour is swept by a carrier gas stream (typically Ar) into the ICP (Fig. 1) for further atomization and excitation/ionization. This is in marked contrast to electrothermal atomic absorption spectrometry (ETAAS) using a graphite furnace in which the ETV device (ie. the furnace) must generate an atomic population as a precursor to atomic absorption. Three key advantages are realized by coupling ETV devices to ICP atomic emission spectrometers. Firstly the matrix effects in ETV-ICP-AES are not as significant as in ETV-AAS because the ETV device simply vaporizes the analyte species whereas in ETV-AAS it must generate free analyte atoms.Secondly vaporization (ETV device) is separate from atomization and excitation/ionization (ICP). The separate control of the ETV and the ICP facilitates independent optimization and results in a system with an overall improved analytical capability. Thirdly when coupled to a polychromator ICP- AES system simultaneous multi-element determinztions can be made. There are also advantages to using ETV devices with ICP-MS. For example the lack of continuous sample aspiration into the plasma eliminates water vapour and results in 'dry' p1asmas.35~42.50~51 As a consequence reductions in spectrosc~pic~~ (i. e. spectral overlaps arising from polyatomic oxide and hydroxide species) and non- * Present address Department of Chemistry University of t To whom correspondence should be addressed.Waterloo Waterloo Ontario N2L 3G1 Canada. Entrance Refractor Concave slit plate grating Plasma box \ / \ Data control I :! transDort - Graphite pellet electronics Sample space Fig. 1 Pellet-ETV-ICP-AES system spectroscopics1 (e.g. matrix induced signal changes) inter- ferences are observed. In spite of the advantages direct solid analysis by ETV is not without shortcomings. These are partly a result of the inherent difficulties associated with finding appropriate calibration standards and with sample handling weighing and the assurance of homogeneity. The last point is an important consideration in the analysis of solid .samples. For example the 0.1-5 mg sample mass typically used with solid sample ETV deviceslo1J02 is difficult to handle and because of ~ampling'~~JO~ and homogeneitylo3-lo6 considera- tions might not be representative of the composition of the bulk of the sample.This last issue is addressed in this work by mixing 25 mg of a powdered sample with 225 mg of graphite and forming a 0.25 g pellet. Work in this laboratory with a pellet-DSI device,52 in which a pellet is directly inserted into the ICP demonstrated that the homogeneity problems were reduced and the detection limits significantly improved by the larger sample mass allowed by this system. The larger sample mass is also easier to handle. These pellet-DSI system concepts are directly transferable to the pellet-ETV system used in this work.528 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 1991 VOL.6 In the proposed system a pellet is placed between two electrodes (Fig. 1). The current applied between them causes ohmic heating of the pellet and results in analyte vaporization. Because the pellet also serves as a sample container the need for a sample holder is eliminated. In addition the direct contact of the pellet with the electrodes and the intimate contact of the heated graphite with the sample allows large amounts of sample to be heated rapidly. In this first report a pellet-ETV sample introduction system for ICP-AES is described. This unoptimized proto- type sample introduction system shows considerable prom- ise for the analysis of inorganic powders and for the rapid screening of botanical samples of environmental concern. Experimental Instrumentation The basic components of the ETV-ICP-AES system devel- oped in this work are shown in block diagram form in Fig.1. A list of instrumentation and materials suppliers is provided in Table 1. Basic instrument specifications and typical operating conditions are given in Table 2. Table 1 List of suppliers of instrumentation and materials ICP Spectrometer Readout electronics Acquisition and control ETV and software microcomputer Mixedmill Pellet press Graphite powder Oxide standards Botanical samples Plasma-Them Kresson NJ USA Jarrell-Ash Model 90750 Thermo Jarrell-Ash Franklin MA USA Technical Service Laboratories Missisauga Ontario Canada AST X-former AT 10 MHz 80286 CPU AST Research Irvine CA USA Model HGA 2200 Perkin-Elmer Norwalk CT USA Model 5 100 Spex Industries Metuchen NJ USA Parr Instrument Moline IL USA Bay Carbon Bay City MI USA J.T. Baker Phillipsburg NJ USA Fisher Scientific Fair Lawn NJ USA Aldrich Milwaukee WI USA Ontario Ministry of the Environment Inorganic Trace Contaminants Section Rexdale Ontario Canada Table 2 Instrument specifications and typical operating conditions ICP R.f. generator Frequency Maximum power output Torch Typical power Outer (coolant) gas Intermediate (auxiliary) gas Central (nebulizer) gas Observation height Pol ychromator Focal length Grating Slits Detector Plasma Therm Model 2500 27.12 MHz crystal controlled 2.00 kW Fassel-t ype 1.00 kW forward < 10 W reflected Ar 14 1 min-I Ar 0.5 1 min-' Ar varied from 0.2 to 1.0 1 min-I 15 mm above the load coil Jarrell-Ash Model 90750 0.75 m 2400 grooves mm-I 25 ,um entrance 50 and 100 ,urn exit PMT Readout electronics Technical Service Laboratories (Table 1 ) Observation time 16 s Number of data points 200 Temperature*/"C Ramp/s Hold/s ETV device- Drying cycle Oxide standards 100 tl 2 Botanical samples 300 t l 60 Oxide standards 200 tl 1 Botanical samples 400 1 120 Oxide standards 1250 1 5 Botanical samples 1250 1 5 Charring cycle Vaporization cycle * Nominal values as shown on the Perkin-Elmer controller not measured.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1 VOL.6 529 0 2 4 6 8 10 12 14 16 Time/s Fig. 2 Typical analyte emission temporal behaviour of 250 ng of Pb (obtained running a single element oxide standard) Electrothermal vaporization produces a plug of analyte vapour which when introduced into the plasma generates a transient emission signal.A typical example of analyte emission temporal behaviour is shown in Fig. 2 for 250 ng of Pb. The transient nature of the emission signals (Fig. 2) necessitates the use of a direct reading spectrometer for simultaneous multi-element analyses. This type of signal also dictates the nature of the analogue electronics as the analyte emission temporal behaviour must be monitored as a function of time. However the integrating readout electronics found in most commercially available direct reading ICP spectrometers are sub-optimal for the acquisi- tion of transient signal^.^^-^^ Furthermore because digitized analyte emission temporal behaviour must be examined by the operator during experimentation it determines the software requirements for data acquisition and for signal display and pro~essing.~~ The multichannel photomultiplier tube (PMT) based direct reading spectrometer (Table 2) used for this work was modified by Technical Service Laboratories in order to handle transient signals. 107~108 Briefly the current output of each PMT channel is integrated (for a programmed period of time typically a few ms) using a high-speed operational amplifier with a 100 pF capacitor in its feedback loop.The resultant voltage is digitized using a 12-bit analogue-to- digital converter interfacedlo7 to an IBM PC compatible microcomputer.109 The software allows one channel to be interrogated repeatedly with a minimum conversion time of 50 ps per data point. A number of converted values are then summed to build the dynamic range.The software also permits repeated sequential interrogation of up to 50 channels with a minimum conversion rate of 2 ms per data point per channel. The spectrometer was a direct reading instrument origi- nally designed for an ax. spark source. The direct reader was modifiedlo7 by adding a computer-controlled galvani- cally driven refractor plate behind the entrance slit (Fig. 1). This modification allows rapid measurement (i.e. 10 ms per data point per channel) of the emission signal on and off the spectral peak. In this way a quasi-simultaneous observation of analyte emission temporal behaviour (e.g. on-peak measurement) and background (off-peak measure- ment) can be obtained. This is an important consideration especially for samples with complex matrices.52 The heart of the sample introduction system shown in Fig.1 is the ETV device. This is a commercially available ETAAS system (Table l) which has been slightly modified to accommodate rapid interconversion between an ETAAS configuration and a pellet-ETV-ICP system. The furnace was modified (Fig. 3) by removing the graphite tube and by replacing the graphite contact rings and the left and right observation windows with machined brass blocks (identi- Water k 3 . 0 c m i Water Base Fig. 3 Cross-sectional view of the pellet-ETV device (illustration to scale) Fig. 4 Sample holder (dimensions in cm illustration to scale) fied as brass electrodes in Fig. 3). The modified furnace assembly is enclosed within a Pyrex chamber. The chamber is a hollow cylinder (3.4 cm in length and 2.7 cm in diameter) with two tube connections.One tube serves as a carrier gas inlet and the other as an outlet. Vaporized samples are routed into the plasma using Tygon tubing ( ~ 4 5 cm in length = 5 mm i.d. and x 7 mm 0.d.). As this pilot study was aimed at testing the validity of the pellet-ETV concept the design parameters of the ETV device (i.e. chamber volume and geometry and tube length) were not optimized. Central to the ETV device is the sample holder (Fig. 4). This is a notched coarse-threaded brass screw. By using a pair of tweezers a pelletized sample is placed at the notch located at one end of the sample holder (Fig. 4). At the other end there is an O-ring to prevent vapour leakage.The sample holder is positioned inside the brass electrode as shown in Fig. 3. Electrical contact between the electrodes is achieved through the pellet (Fig. 3). Drying charring and vaporization are obtained by programming the time and the intensity of the current applied to the pellet. During vaporization the pelletized sample is heated to incandescence. Method Sample preparation consisted of grinding botanical samples manually in a mortar for about 15 min (the average particle size was about 160 pm) and mixing the powdered sample with spectroscopically pure graphite (1 + 9). Powdered oxide standards were prepared by dilution with spectros- copically pure graphite according to a procedure described previou~ly.~~ A typical sample preparation time of 15-20 min was required for batch preparation of 5 pellets.An accurately weighed portion of this mixture (250 mg) was hand pressed using a Parr press (Table 1) to form a pellet which is 9 mm long and 4 mm in diameter. Pellets placed in5 30 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 1991 VOL. 6 the ETV device were ramped through a short drying cycle followed by a charring cycle. The temperature and time for these cycles were sample- type dependent (Table 2). These steps were necessary even for the oxide standards as a rapid release of gases often resulted in cracking of the pellets. This in turn gave rise to irregular peak shapes and double or split peaks. With the botanical samples higher temperatures and longer charring times were necessary (Table 2) otherwise the cracks in the pellets were more profound (verified by visual inspection) and highly irregular peaks were observed.Some pellets would even break causing the runs to be aborted. At the end of the charring cycle the temperature was rapidly raised to 1250 "C. Although direct comparisons with other ETV sample introduction systems might be misleading owing to differences in sample-holder design (2. e. graphite-cup -rod and -tube) the vaporization tem- peratures used with these systems are typically higher (ie. 1600-2400 0C,23 2400 0C,10J7 2500 0C,33 2600 0C,11329 2700-3000 0C,22 and 3000 0C.32 At 1250 "C very little graphite was released from the pellet and the ETV device could be used continuously for about half a day without any noticeable deposits or memory effects.The use of higher vaporization temperatures was also tested and resulted in sharper peaks. However at about 1400 "C some graphite was also released this was visually con- firmed by the observation of a weak carbon emission in the plasma and a carbon deposit in the chamber and the Tygon tube. As the temperature was raised from 1400 to 1800 "C a progressively more intense carbon emission and a larger deposit were observed. Larger deposits caused increasingly persistent memory effects thus requiring a more frequent cleaning of the chamber and the tube. This disturbed the operation and increased the analysis time. At about 1800 "C the chamber and the tube had to be cleaned after every run. Although such temperatures can be advantageous in overcoming matrix effects they also give rise to a pressure pulse (as explained later).Single channel data acquisition parameters were chosen so that 200 points were acquired using a data acquisition rate of about 13 data points per second. From the results shown in Fig. 2 it can be concluded that adequate data were acquired to give a suitable peak shape. Data were manipu- lated on the IBM PC compatible microcomputer109 using a spreadsheet and were transferred to an Apple Macintosh microcomputer for further processing display and presentation. Results and Discussion Analyte Emission Temporal Behaviour In order to establish the nature of the signals generated by this system analyte emission temporal behaviour was recorded for several elements. Results obtained by running single element oxide standards are shown in Fig.5 for 250 ng of As Cd Pb and Zn and 25 ng of Hg and Mn. A peak appearance time (defined as the time between the start of the vaporization cycle and the peak maximum) of less than 10 s was observed for all of the elements tested. Typical examples of the sequence in which the analytes volatilized from the pellet (ie. a 'volatility sequence') are shown in Fig. 6. These were obtained using single element oxide standards (Table 1) and the conditions listed in Table 2. In general peak widths (at half-height) ranged between 1 and 2.5 s (Figs. 5 and 6) and the post-vaporization level was of approximately equal magnitude to that of pre-vaporization. The data also show that relatively high concentrations of elemental impurities are present in commercially available spectroscopic graphite (Fig.5 lower trace). The absence of a pressure pulse of the type observed 150000 100000 50000 0 h . 25000 7 2oooo f 10000 .= 15000 x c .- cn 5000 a Y - c o 30000 20000 10000 0' ' ' ' ' ' ' ' 1 60000 50000 40000 30000 20000 10000 0 4000 3000 - 2000 - 1000 - - - 0 2 4 6 8 I01214 16 0 2 4 6 8 10121416 Timels Fig. 5 Typical analyte emission temporal behaviour (obtained running single element oxide standards). Upper trace analyte emission (250 ng) of As Cd Pb and Zn; and analyte emission (25 ng) of Hg and Mn. Lower trace graphite blank 100 1 25 - t= I.. I Mn Zn Pb 0 2 4 6 8 10 12 14 16 Time/s Fig. 6 Volatility sequence for Hg Mn Zn and Pb (see text for discussion) during the high temperature vaporization stage in other ETV s ~ s ~ ~ ~ s ~ ~ ~ J ~ J ~ is noteworthy.The pulse mani- fests itself as a decrease in the intensity of the plasma background and has been attributed to the rapid heating of the carrier gas by the sample holder. The rapid heating induces gas expansion which creates a momentary increase in the carrier gas flow rate and results in a corresponding decrease in plasma continuum emission. An increase in the length of the Tygon tube,I0 sample carrier gas flow rate13 and observation height,13J5 a reduction in the volume of the ETV chamber1* and a double wall glass chamber15 have been identified as means by which the adverse effects of background signal depression are reduced. The lack of an intense pressure pulse with this system was attributed to the use of low vaporization temperatures and the small surface area of the pellet. However with higher vaporization temperatures a pressure pulse became noticeable with this system.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1.VOL. 6 53 1 0 0.2 0.4 0.6 0.8 1 .o 1.2 Carrier gas flow rate/! min-’ Fig. 7 Effect of central tube flow rate on signal-to-background ratios (obtained running single element oxide standards) of 250 ng of Pb and Zn; and 25 ng of Mn 2 .- 2 - 6000 5000 ‘ 4000 3000 2000 1000 0 v) - .4 /I I /I f? 0 2 4 0 2 4 0 2 4 Time/s Fig. 8 Reproducibility (obtained running single element oxide standards) for 250 ng of Pb and Zn (see text for discussion) Operating Conditions Although the data shown in Figs. 5 and 6 illustrate the basic temporal characteristics of the ETV-ICP-AES signals the exact nature depends on a number of system parameters.These include vaporization temperature and duration tube length chamber volume and geometry plasma power and gas flow rates. As a proof-of-concept approach was adopted in this study no attempt was made to examine the effect of all of these parameters on the analyte emission temporal behaviour. Only the effect of the gas flow rate in the central tube (which is equivalent to the nebulizer flow rate of pneumatic nebulizer systems) was studied. The gas flow rate in the central tube was varied from 0.2 to 1.0 1 min-l. The results (reported as peak heights) obtained by running the single element oxide standards (250 ng of Pb and Zn and 25 ng of Mn) are shown in Fig.7. It is apparent that this gas flow rate is a key parameter and that compromise conditions are important with this system. Although not shown ti 2 gas flow rate in the central tube also had an effect on peak shapes. In general higher flow rates produced sharper peaks as expected. Basic Analytical Performance Characteristics In order to obtain an indication of the potential analytical performance characteristics of this system the precision was measured the detection limits were determined and calibration graphs constructed. Unless otherwise stated the parameters listed in Table 2 and pellets containing single element oxide standards were used. Fig. 8 shows the signals obtained for three successive runs of 250 ng of Pb and Zn. In general peak heights and peak shapes were reproducible. Similar results were obtained for Cd and Mn.Average relative standard deviations deter- mined from six replicate runs were 4.8% (peak height) and 7.0% (peak area). These results compare favourably with those reported by other workers for solid-sample-ETV-ICP systems2-6 but are much better than those for a pellet-DSI- ICP In contrast to the reproducible signals obtained when running pellets containing the same amount of analyte the shape of the emission signal changes considerably as a function of the amount of analyte in a pellet. For example signals for 2500 and 25 ng of Cd are shown in Fig. 9. Similar changes in analyte emission temporal behaviour have been reported by other w o r k e r ~ . ~ ~ J ~ ~ These are analogous to the ‘concentration effects’ observed in ETAAS.111J12 It could be suggested that the changes are due to thermal vaporization effects to diffusion and transport effects (Cd transport efficiency has been found to increase with sample mad2) and to changes in plasma characteristics. However more work remains to be done so as to establish the influence these parameters exert on peak shapes.Detection limits (Table 3) determined using peak height measurements are three times the standard deviation of the background while running 25 ng of a single element oxide standard. Also included in Table 3 are the detection limits obtained when running liquid samples using a conventional pneumatic nebulizer and the detection limits reportedS2 using a pellet-DSI device. The experimental configuration used for the determination of the detection limits remained unchanged with the exception of the sample introduction system.In general the detection limits obtained using a pellet-ETV system are 6-10 times better than those 300000 r 1 200000 100000 h c v1 c 3 .- .- c. g o n v 2 2000 0 2 4 6 8 10 12 14 16 Ti m e/s Fig. 9 Effect of concentration on peak shapes for (a) 2500 ng of Cd and (b) 25 ng of Cd. Lower trace graphite blank532 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1 VOL. 6 Table 3 Detection limits Pellet-ETV Line/ Pellet-DSI* Liquid Element nm ppb pg (PPb) (PPb) C d I 228.8 1 300 10 26 Pb I 220.3 0.6 150 70 28 Zn I1 213.8 3 800 20 82 *See reference 52. Table 4 Analysis of MOE botanical samples (concentration in PPm) Pb Zn Sample type Found MOE Found MOE V85-1 l o + 1 19k2.0 370k 12 140-r- 11 White birch 3.2 20.2 4 370k5 200 Norway maple 39 2 1 95 39k2 39 reported for a pellet-DSI-ICP-AES systems2 and 30-50 times better than those obtained when running liquid samples.Linear calibration graphs for Cd Pb and Zn covering a concentration range of 2-3 orders of magnitude were established using peak heights and single element oxide standards. Non-linear calibration graphs were obtained for As Hg and Mn. No explanation can be offered at this time. In order to evaluate the analytical performance of the pellet-ETV device further and to test the feasibility for the direct analysis of ‘real’ samples botanical samples provided by the Ontario Ministry of the Environment (MOE) were used. Included in these are sample types designated as V85- 1 Norway maple and White birch by the MOE.Certified values and with the exception of V85-1 statistical data are not available to us. Elemental concentrations quoted by the MOE and reported in Table 4 are the average of a large number of analyses of these sample types. According to the MOE these concentrations were obtained after a 3 h open- vessel hot-plate digestion and analysis by ICP-AES. In this laboratory sample preparation consisted of mixing 0.300 g of a ground botanical reference sample with 2.700 g of spectroscopic graphite as previously described for batch preparation of 12 pellets. The results obtained using the calibration graphs constructed from the single element oxide standards are shown in Table 4 and are encouraging considering the sample size and the fact that there was neither matrix matching nor sample pre-treatment.Conclusion The results of this preliminary investigation show that a pellet-ETV-ICP-AES system is a promising method for elemental determinations with powdered samples. This system is suitable for the rapid screening of powdered botanical samples of environmental concern. Improve- ments in the analytical performance characteristics of this system are expected by designing an optimized pellet-ETV device and by using a spectrometer configured for the ICP. Further improvements are expected by coupling a pellet- ETV device to an ICP mass spectrometer. For example if the improvement in detection limits is similar to that observed for the pellet-DSI-ICP-MS,s2 detection limits for the pellet-ETV-ICP-MS should also improve by as much as three orders of magnitude.In addition owing to dry sample introduction afforded by ETV devices significant reduc- tions in spectroscopic and non-spectroscopic effects are expected (similar to DSI-ICP-MSSo~sl). Financial assistance from the Ontario Ministry of the Environment (Projects 414G and 270R) and L. Blain’s assistance are gratefully acknowledged. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 References Kantor T. Spectrochim. Acta Part B 1983 38 1483. Ng K. C. and Caruso J. A. Appl. Spectrosc. 1 985 39 7 19. Matusiewicz H. J. Anal. At. Spectrom. 1986 1 171. Broekaert J. A. C. and Boumans P. W. J. M. in Inductively Coupled Plasma Emission Spectrometry ed.Boumans P. W. J. M. Wiley New York 1987 part I ch. 6 p. 296. Van Loon J. C. Haraguchi H. and Fuwa K. in Inductively Coupled Plasma Emission Spectrometry ed. Boumans P. W. J. M. Wiley New York 1987 part 11 ch. 7 p. 217. Routh M. W. and Tikkanen M. W. in Inductively Coupled Plasmas in Analytical Atomic Spectrometry eds. Montaser A. and Golightly D. W. VCH New York 1987 ch. 12 p. 431. Broekaert J. A. C. Leis F. Raeymaekers B. and Zaray Gy. Spectrochim. Actaj Part B 1988 43 339. Littlejohn D. Jowitt R. Shuttler I. L. Sparkes S. T. Tyson J. F. and Walton S. J. J. Anal. At. Spectrom. 1990,5 179R. Nixon D. E. Fassel V. A. and Kniseley R. N. Anal. Chem. 1974 46 210. Gunn A. M. Millard D. L. and Kirkbright G. F. Analyst 1978 103 1066. Kirkbright G.F. and Snook R. D. Anal. Chem. 1979 51 1938. Millard D. L. Shan H. C. and Kirkbright G. F. Analyst 1980 105 502. Ng K. C. and Caruso J. A. Anal. Chim. Acta 1982 143 209. Ng K. C. and Caruso J. A. Anal. Chem. 1983,55 2032. Barnes R. M. and Fodor P. Spectrochim. Acta Part B 1983 38 1191. Matusiewicz H. and Barnes R. M. Spectrochim. Acta Part B 1984 39 891. Swaidan H. M. and Christian G. D. Anal. Chem. 1984,56 120. Matusiewicz H. and Barnes R. M. Spectrochim. Acta Part B 1985,40 29. Matusiewicz H. and Barnes R. M. Spectrochim. Acta Part B 1985,40 41. Long S. E. Snook R. D. and Browner R. F. Spectrochim. Acta Part B 1985 40 553. Matusiewicz H. Fricke F. L. and Barnes R. M. J. Anal. At. Spectrom. 1986 1 203. Kumamaru T. Okamoto Y. and Matsuo H. Appl. Spec- trosc.1987 41 918. Alvarado J. Cavalli P. Omenetto N. Rossi G. and Ottaway J. M. J. Anal At. Spectrom. 1987 2 357. Isoyama H. Okuyama S. Uchida T. Takeuchi M. Iida C. and Nakagawa G. Anal. Sci. 1990 6 555. Hull D. R. and Horlick G. Spectrochim. Acta Part B 1984 39 843. Blakemore W. M. Casey P. H. and Collie W. R. Anal. Chem. 1984,56 1376. Ohls K. and Hutsch B. ICP InJ Newsl. 1986 12 170. Ohls K. Dewies J. and Loepp H. ICPlnJ Newsl. 1986,12 177. van Berkel W. W. and Maessen F. J. M. J. Spectrochim. Acta Part B 1988 43 1337. Reisch M. Nickel H. and Mazurkiewicz M. Spectrochim. Acta Part B 1989 44 307. Nickel H. Reisch M. and Mazurkiewicz M. Fresenius 2. Anal. Chem. 1989 335 631. Ida I. Yoshikawa H. Ishibashi Y. and Gunji N. Anal. Sci. 1989 5 615. Atsuya I. Itoh T.Kurotaki T. Spectrochim. Acta Part B 1991,46 103. Gray A. L. and Date A. R. Analyst 1983 108 1033. Gray A. L. Spectrochim. Acta Part B 1986 41 15 1. Date A. R. and Cheung Y. Y. Analyst 1987 112 1531. Park C. J. van Loon J. C. Arrowsmith P. and French J. B. Can. J. Spectrosc. 1987 32 29. Park C. J. van Loon J. C. Arrowsmith P. and French J. B. Anal. Chem. 1987 59 2191. Park C. J. and Hall G. E. M. J. Anal. At. Spectrom. 1987,2 473.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1 VOL. 6 533 40 Park C. J. and Hall G. E. M. J. Anal. At. Spectrum. 1988,3 355. 41 Hall G. E. M. Pelchat J.-C. Boomer D. W. and Powell M. J. Anal. At. Spectrum. 1988 3 791. 42 Gregoire D. C. J. Anal. At. Spectrum. 1988 3 309. 43 Gregoire D. C. Anal. Chem. 1990 62 141. 44 Tsukahara R.and Kubota M. Spectrochim. Acta Part B 1990 45 779. 45 Shen W.-L. Caruso J. A. Fricke F. L. and Satzger R. D. J. Anal. At. Spectrom. 1990 5 45 1. 46 Shibata N. Fudagawa N. and Kubota M. Anal. Chem. 1991 63 636. 47 Karanassios V. and Horlick G. Spectrochim. Acta Rev. 1990 13 89. 48 Karanassios V. and Horlick G. Spectrochim. Acta Part B 1990 45 85. 49 Karanassios V. and Horlick G. Spectrochim. Acta Part B 1989,44 1345. 50 Karanassios V. and Horlick G. Spectrochim. Acta Part B 1989 44 1361. 51 Karanassios V. and Horlick G. Spectrochim. Acta Part B 1989,44 1387. 52 Blain L. Salin E. D. and Boomer D. W. J. Anal. At. Spectrom. 1989 4 721. 53 Ohls K. and Sommer D. Fresenius Z. Anal. Chem.. 1979 296 241. 54 Keilsohn J. P. Deutsch R. D. and Hieftje G. M.. Appl.Spectrosc. 1983 37 10 1. 55 Farnsworth P. B. and Hieftje G. M. Anal. Chem. 1983,55 1414. 56 Jiang S. J. and Houk R. S. Anal. Chem. 1986 58 1739. 57 Human H. G. C. Scott R. H. Oakes A. R. and West C. D. Analyst 1976 101 265. 58 Scott R. H. Spectrochim Acta Part B 1978 33 123. 59 Marks J. Y. Fornwalt D. E. and Yungk R. E. Spectrochim. Acta Part B 1983 38 107. 60 Beaty J. S. and Belmore R. J. J. Test. Eval. 1984 12 212. 61 Aziz A. Broekaert J. A. C. Laqua K. and Leis F. Spectrochim. Acta Part B 1984 39 1091. 62 Beckwith P. M. Mullins R. L. and Coleman D. M. Anal. Chem. 1987 59 163. 63 Lemarchand A. Labarraque G. Masson P. and Broekaert J. A. C. J. Anal. At. Spectrom. 1987 2 481. 64 Ono A. Saeki M. and Chiba K. Appl. Spectrosc. 1987 41 970. 65 Prell L. J. and Koirtyohann S.R. Appl. Spectrosc. 1988,42 1221. 66 Raeymaekers B. Van Espen P. Adams F. and Broekaert J. A. C. Appl. Spectrosc. 1988 42 142. 67 Thompson M. Goulter J. E. and Sieper F. Analyst 1981 106 32. 68 Carr J. W. and Horlick G. Spectrochim. Acta Part B 1982 37 1. 69 Ishizuka T. and Uwamino Y. Spectrochim. Acta Part B 1983 38 5 19. 70 Dittrich K. and Wennrich R. Prog. Anal. At. Spectrosc. 1984 7 139. 71 Gray A. L. Analyst 1985 110 551. 72 Arrowsmith P. Anal. Chem. 1987 59 1437. 73 Tremblay M. E. Smith B. W. Leong M. B. and Wineford- ner J. D. Spectrosc. Lett. 1987 20 31 1. 74 Cremers D. A. and Radziemski L. J. in Laser Spectroscopy and its Applicatjons eds. Radziemski L. J. Solarz R. W. and Paisner J. A. Marcel Dekker New York 1987 ch. 5. 75 Dittrich K. Niebergall K.and Wennrich R. Fresenius 2. Anal. Chem. 1987 328 330. 76 Arrowsmith P. and Hughes S. K. Appl. Spectrosc. 1988,42 1231. 77 Su G. and Lin S. J. Anal. At. Spectrom. 1988 3 841. 78 Hager J. W. Anal. Chem. 1989 61 1243. 79 Thompson M. Chenery S. and Brett L. J. Anal. At. Spectrom. 1989 4 11. 80 Lin S. and Peng C. J. Anal. At. Spectrom. 1990 5 509. 81 Richner P. Borer M. W. Brushwyler K. R. and Hieftje G. M. Appl. Spectrosc. 1990 44 1290. 82 Imai N. Anal. Chim. Acta 1990 235 381. 83 Pang H. Wiederin D. R. Houk R. S. and Young E. S. Anal. Chem. 1991 63 390. 84 Denoyer E. R. Fredeen K. J. and Hager J. W. Anal. Chem. l991,63,445A. 85 Trivedi K. M. Brewer S. W. Jr. and Sacks R. D. Appl. Spectrosc. 1990 44 367. 86 Hoare H. C. and Mostyn R. A. Anal. Chem. 1967 39 1153.87 Dagnall R. M. Smith D. J. West T. S. and Greenfield S. Anal. Chim. Acta 1971 54 397. 88 Ng K. C. Zerezghi M. and Caruso J. A. Anal. Chem. 1984 56 4 17. 89 De Silva K. N. and Guevremont R. Spectrochim. Acta Part B 1986 41 865. 90 Guevremont R. and De Silva K. N. Spectrochim. Acta Part B 1986 41 875. 91 Allen G. M. and Coleman D. M. Appl. Spectrosc. 1987,41 381. 92 Pfannerstill P. E. Caruso J. A. and Willeke K. Appl. Spectrosc. 1989 43 626. 93 Pfannerstill P. E. Creed J. T. Davidson T. M. Caruso J. A. and Willeke K. J. Anal. At. Spectrom. 1990 5 285. 94 De Silva K. N. and Guevremont R. Spectrochim. Acta Part B 1990 45 997. 95 De Silva K. N. and Guevremont R. Spectrochim. Acta Part B 1990 45 1013. 96 Guevremont R. and De Silva K. N. Spectrochim. Acta Part B 1991 46 67.97 Mohamed N. and Fry R. C. Anal. Chem. 1981,53,450. 98 McCurdy D. L. and Fry R. C. Anal. Chem. 1986,58,3126. 99 Ebdon L. and Wilkinson J. R. J. Anal. At. Spectrom. 1987 2 325. 100 Ebdon L. Foulkes M. E. and Hill S. J. Anal. At. Spectrom. 1990 5 67. 101 Langmyhr F. J. Fresenius 2. Anal. Chem. 1987 328 315. 102 Bradshaw D. and Slavin W. Spectrochim. Acta Part B 1989 49 1245. 103 Kratochvil B. and Taylor J. K. Anal. Chem. 1981 53 924A. 104 Kratochvil B. Wallace D. and Taylor J. K. Anal. Chem. 1984 56 113R. 105 Ingamells C. O. Talanta 1974 21 141. 106 De Silva K. N. and Guevremont R. Spectrochim. Acta Part B 1990 45 933. 107 Legere G. and Burgener P. ICP ZnJ Newsl. 1988 13 521. 108 Nygaard D. D. ICP InJ Newsl. 1988 13 735. 109 Karanassios V. and Horlick G. Talanta 1985 32 615. 110 Tikkanen M. W. and Niemczyk T. M. Anal. Chem. 1986 58 366. 111 McNally J. and Holcombe J. A. Anal. Chem. 1987 59 1105. 112 Harnly J. M. J. Anal. At. Spectrom. 1988 3 43. Paper l/UO6 72J Received February 13th 1991 Accepted May 30th I991